Industry data from South Africa's major retirement fund administrators shows strikingly high rates of repeat Two-Pot Savings Pot withdrawals: Momentum reported 95% of 2027 tax year claimants were repeat users, Sanlam reported 75%, and Alexforbes reported 67%. SARS itself reported R79.3 billion approved for withdrawal since the September 2024 launch, with 5.6 million people applying for tax directives.
This pattern, sometimes described as a "13th cheque" habit, diverges from the system's original design as an emergency safety valve, a related but distinct trend also shows half of South Africans cashing out retirement savings entirely when changing jobs. Most claimants understand the retirement and tax implications, average claim sizes have actually been declining, suggesting smaller, more frequent liquidity needs rather than large one-off emergencies.
Two-Pot Repeat Withdrawals: The Numbers
Figures vary somewhat by fund administrator and reporting period, treat these as representative industry data rather than a single official total.
SARS reported that between the system's launch on 1 September 2024 and 28 February 2026, R79.3 billion had been approved for withdrawal from Savings Pots, with a resulting tax liability of R21.4 billion. A total of 5.6 million people applied for the required tax directives over that period, a genuinely high uptake relative to South Africa's total working population.
This confirms the Two-Pot Retirement System has been used at meaningful scale since its introduction, the question this article focuses on specifically is HOW it's being used, and whether that matches the system's original intent.
Data from major fund administrators consistently shows high rates of repeat withdrawal among members who've claimed before. Momentum reported that 95% of members withdrawing in the 2027 tax year were doing so for a second or third time. Sanlam's equivalent figure was 75% repeat withdrawals in the 2026 tax year. Alexforbes found 67% of members who claimed in the 2025 tax year also submitted a claim again in 2026, with 38% of 2026 claimants already claiming again within the first month of the 2027 tax year.
| Administrator | Repeat Rate | Period |
|---|---|---|
| Momentum | 95% | 2027 tax year |
| Sanlam | 75% | 2026 tax year |
| Alexforbes | 67% | 2025โ2026 tax years |
The Two-Pot system was structured as a deliberate compromise between complete preservation of retirement savings and unrestricted access, intended to function as a genuine safety valve for real financial emergencies, not as a predictable annual spending account. The high, consistent repeat withdrawal rate suggests that for a meaningful portion of members, the Savings Pot has instead become something closer to an expected, recurring source of supplementary income, informally described by some commentators as a "13th cheque."
This behavioural shift matters structurally: a system designed around occasional, emergency access behaves very differently, both for individual retirement outcomes and for fund administration, when access becomes a routine, expected annual event for a large share of the member base.
Industry reporting indicates the Savings Pot is primarily being used for short-term liquidity needs, debt repayment, school fees, and medical expenses feature prominently, connecting to the broader pattern covered in our look at South Africa's household debt levels, rather than large, one-off emergencies. This is reflected in the claim size data: the average claim value has been declining over time, from roughly R12,666 to R14,250 at launch in September 2024, down to approximately R9,290 by March 2026, with 71% of claims now falling under R10,000.
This declining average claim size, combined with the high repeat rate, paints a fairly clear picture: many members are drawing smaller amounts more frequently for ongoing liquidity pressure, rather than the occasional larger emergency withdrawal the system's design more naturally anticipated.
Survey data from Alexforbes found that 96% of claimants understood their withdrawal would reduce their eventual retirement savings, and an equal 96% were aware of the tax implications involved, this isn't fundamentally a knowledge gap. What appears to genuinely surprise many members is the practical SIZE of the deduction: withdrawals are taxed at your full marginal PAYE rate with no tax-free portion at all, unlike the more favourable lump sum tax table that applies at actual retirement.
Any arrear tax debt owed to SARS is also deducted first, before the remaining amount is paid out, meaning the final payout received is often considerably smaller than the headline withdrawal amount a member initially expected, even when they understood the general principle beforehand.
At the individual level, repeated withdrawal genuinely compounds the opportunity cost to eventual retirement outcomes, each withdrawal removes capital that would otherwise have continued growing tax-efficiently until retirement, and repeated withdrawals compound this lost growth significantly over a working lifetime, similar in principle to why how much you risk trading matters so much for long-term outcomes. This is a real, meaningful cost even though awareness of the general principle is high.
At the system-wide level, the picture is more contained than some initial concerns suggested, Momentum reported that while 33.7% of members have made at least one withdrawal, the cumulative impact on total fund assets remains around 1.97%. This suggests the repeat-withdrawal pattern, while a genuine concern for the specific individuals affected, hasn't yet meaningfully undermined the broader retirement savings system at a national scale.
SARS reported that as of late February 2026, R79.3 billion had been approved for withdrawal from savings pots since the system's September 2024 launch, with a tax liability of R21.4 billion, and 5.6 million people had applied for tax directives. This represents a very high uptake relative to initial expectations for the reform.
Industry data varies by fund administrator but consistently shows high repeat usage: Momentum reported 95% of members withdrawing in the 2027 tax year were doing so for a second or third time, while Sanlam reported 75% repeat withdrawals in the 2026 tax year, and Alexforbes found 67% of 2025 tax year claimants also claimed again in 2026.
Not entirely. The Two-Pot system was designed as a compromise between complete preservation and unrestricted access, intended to provide a genuine safety valve, not an annual spending account. The high rate of repeat, sometimes described as a "13th cheque" pattern, suggests many members are using it as recurring supplementary income rather than reserving it for genuine emergencies.
Industry reporting indicates the Savings Pot is primarily being used for short-term liquidity needs including debt repayment, school fees, and medical expenses, rather than one-off emergencies. The average claim value has also been declining over time, from around R12,666 to R14,250 at launch down to roughly R9,290 by March 2026, with 71% of claims now under R10,000.
Survey data from Alexforbes found 96% of claimants understood their withdrawal would reduce retirement savings, and an equal 96% were aware of the tax implications, awareness generally isn't the issue. What often surprises members is the SIZE of the tax deduction in practice, since withdrawals are taxed at your marginal PAYE rate with no tax-free portion, and any outstanding SARS debt is deducted first, meaning the actual amount received is often considerably less than the headline withdrawal figure.
The picture is mixed. Momentum reported that while 33.7% of members have made at least one withdrawal, the cumulative impact on total fund assets remains relatively limited, around 1.97%. This suggests that while individual members withdrawing repeatedly face a real compounding cost to their own retirement outcome, the system-wide effect on total retirement savings has been more contained than some initial concerns suggested.
This article draws on official SARS reporting and published industry data from major retirement fund administrators. Always verify current figures directly at each source.